REGIONAL PREDICTION MODELS FOR THE ABOVEGROUND BIOMASS ESTIMATION OF Eucalyptus grandis IN NORTHEASTERN ARGENTINA

Different industrial, environmental and energy interests have led, in recent years, to the analysis of different possibilities for utilization of forest biomass is gaining an increasing role, so that quantification of the biomass from these forest systems is an essential first and an unavoidable step. The aim of this study was to develop models which allow estimate the biomass of different compartments of Eucalyptus grandis trees. To do this 41 trees from stands of different ages were employed, between 4 and 32 years, located in the NE region of Argentina. Twenty three (23) trees were felled in the North and High area of Mission (Zone 1) and eighteen 18 in Southern Misiones and Corrientes NE (Zone 2). The Biomass was determined from leaves, branches less than 5 cm, branches larger than 5 cm, branches total, stem and tree totals biomass. For the models fitting were taken several independent variables, diameter at breast height (dbh), total tree height (h), the product (dap2 * h) and the dummy variable (zone). In the case of the biomass branches and leaves, because the results achieved were not satisfactory with the use of these variables, there were examined models that add the site index, spacing factor, diameter at the base of green crown (dbgc), height to green crown base (hgcb), length of green crown (lgc), the cross-sectional area of sapwood at the base of the green crown (c Sap bgc) and different combinations of these variables. The models were selected based on coefficient of determination and mean square error. For biomass of branches smaller than 5 cm, branches larger than 5 cm, branches total, stem and total biomass of the tree, the best performing models were those which incorporated as explanatory variables, the diameter at breast height, the total tree height and the "dummy" variable (zone). For biomass of leaves the variables age, basal area and number of trees per hectare improved the estimate. The coefficients of determination for the model of the total tree and stem biomass was 0.99, for the total of branches was 0.83 for branches smaller than 5 cm 0.69, for branches larger than 5 cm gave 0.53 and was 0.65 for leaves. It is recommended to increase the database for the purpose of trying to improve predictions of biomass of leaves and branches.